top of page

This paper describes an experiment using a small, cheap, portable Aether wind detector that can be made for just a few hundred dollars and can successfully detect the light speed anisotropy in the Earth’s reference frame and determine its magnitude and direction.

Image 1: The full experimental setup (except for the PC that collects the data and the steel enclosure)

Image 2: The stack up of 10 double-sided PCBs that have the looping copper track etched into it. All PCBs are connected in series.

Image 3: A close up view of the etched copper track on the top PCB.

Image 4: The steel enclosure box that houses the PCB stack, Vector Network Analyzer and cables, to prevent external magnetic and RF interference with the recorded results data.

Image 4: The Azimuth/Altitude of the expected Aether wind at 7:12pm on 11/04/2023 in Melbourne, Australia at the location where the experiment was conducted.


Image 5: The Phase timing graph versus frequency graph for the North/South & East/West directions.


Image 7: The Phase timing difference graph versus frequency graph for the North/South versus East/West directions. The Orange line is covered by the Yellow line from the theoretical model (as they have almost identical values) and is the final average value computed from all of the gathered experimental data. The Green line is a 10-sample running average of the data points.

Progress Update

I have upgraded the Phase detector used in the experiment to get better results and greatly improve the signal to noise ratio of the results data. Instead of using a nanoVNA device to scan a range of frequencies, I am now using a tinyPFA device to accurately measure Phase timing differences between Channel 1 and Channel 2 of the device. The tinyPFA device is actually a nanoVNA-H4 device which has been reprogrammed using a firmware image for the tinyPFA functionality. To see how to do this, see this website:


The main difference in the use of the modified device is that both Channel 1 and Channel 2 ports are inputs and the stable clock source used in the experiment must be provided externally from a Signal Generator (or another stable oscillator). For my experiment I am using a 10MHz pure sine wave at 12V from the signal generator. I need the high voltage as the voltage drop across the almost 2km long copper track in my detector is large and I need a sufficient signal to detect at the other end of the long track. This means that I needed to obtain a 20dB attenuator for the signal at the start of the long track, that feeds into Channel 1 of the tinyPFA, otherwise it will be damaged (the signal power must not exceed +10dBm). With this setup I can get about -14dBm in Channel 1 and -26dBm into Channel 2 of the tinyPFA. These levels are sufficient to get a good reading.

After configuring the tinyPFA and connecting it to the free TimeLab software (to display the results), I started recording some data. The initial data was promising – showing a phase difference between orthogonal orientations of the device.

In some further testing, there were some discrepancies in results depending if the device was rotated clockwise or anticlockwise (between the two orthogonal directions). So, further measures were taken to eliminate these effects. The primary cause was found to be due to flexing of the coaxial cable from the signal generator to the device inside the steel box during rotation of the device, which had the effect of changing the recorded phase times. So, the Signal Generator was placed directly on top of the device such that it rotates with the device when its direction is being changed and the coaxial cable connecting it to the device remained stationary. This allowed for a shorter coaxial cable between the SigGen and the device and prevented any flexing of the coaxial cable. I identified another possible issue to do with potential interference with the electronic detector from the Earth’s magnetic field. The steel box provides some shielding from this, but just to make sure the Earth’s magnetic field was eliminated to s sufficiently low level I added a MuMetal lining to the inside of the steel box. MuMetal has a high permeability and is used to block magnetic fields.

Here are some pictures of the MuMetal lining and new testing setup:

After resolving the technical problems with taking the measurements, I obtained the average of 10 consecutive North/South versus East/West phase time difference readings and obtained an average phase time difference of 4.59E-12 seconds. The graph above depicts the phase difference between the two channels of the tinyPFA phase detector (after being zeroed at the start of the experiment). As the travel time of the Electromagnetic wave down the electrical conductor (which is a longitudinal wave rather than a transverse wave which is the case in an optical interferometer) in the transverse direction (to the direction of motion through the Aether) is slightly less than for the light traveling in the longitudinal direction, it would be expected that there is a slightly smaller phase time difference for the signal traveling East/West when compared to the signal traveling North/South. This is what was seen in the results, as you can see on the graph in Fig. 8.

The model (for the date, time and location of the experiment - see Appendix A) predicted a phase time difference of 4.71E-12 seconds. So, in this run of the experiment, the measured phase timing was 97.33% of the prediction from the model (see Appendix B). This corresponds to a measured Aether wind speed of about 479.8km/sec, which matches the previously determined best value from other experiments of 486km/sec at 98.72% of the same value.

bottom of page